|
|
|
|
Good afternoon, My name is David Walker and I wear several hats. I'm the director of content for the online home loan service eChoice. I write the Webmechanic column for The Age and the Sydney Morning Herald. I run a personal Web site called Shorewalker.com. Steve Cassidy told me he wanted me to speak at this conference because I'm an "iconoclast". The original iconoclasts destroyed works of art, in the Byzantine Empire of the 8th and 9th centuries. I trust he didn't expect me to do any of that. I presume he meant I was the sort of person who tries to overthrow popular ideas. I certainly used to do that, back in the days of the dot-com madness. These days people are supposed to have wised up. These days, I hope I might be classified as a "former iconoclast".
I began writing about the Internet five years ago because more and more of what I saw happening in the Internet environment made no sense to me. Specifically, billions of dollars were beginning to disappear into development projects that didn't meet people's needs. Instead, money was being poured into concepts built on bizarre economics, fantasies about media use and over-hyped technology. I went to work for a company - eChoice - which believed that Internet-centred business obeyed all the normal rules of business, media and software. I and my colleagues looked like idiots until about last April. But eChoice is now perhaps the strongest surviving dot-com business in Australia. And there's a sudden interest in creating Web projects that perform real work within a reasonable budget. People are starting to realise the costs of Net hype.
The Web has freed us suddenly and forever from the tyranny of the printing press. Anyone with a $400 PC and an Internet connection can publish thousands of words and images in a simple Web site. But most organisations do not keep their Web sites simple. They expand those sites in size, complexity and scope. Before long, they have ambitions that can only be fulfilled through a relational database and a bunch of program logic. At this point your site ceases to be a brochure and becomes an Internet software development project - and things can get very expensive very fast. At this point people need to sit down and think about whether their large investment will generate large returns. For the past few years, this thinking hasn't been done. Organisations have typically preferred to give in to the hype - all those ads that said they had to be e-enabled tomorrow or else. So they have typically:
Let's start by looking at the glamorous, sexy stuff, the stuff that gets hyped the most - what people get out of online activity.
Here it is. Hmmm … 85 per cent use email, 82 per cent use search engines, 34 per cent do personal health research and read the daily newspaper, 24 per cent check the weather. What the heck? We're in the middle of a cool transformational high-technology new-media revolution here, and people are using it to find out whether they should put on a raincoat? Well, yes. Online research can often be of dubious quality. But on this topic, all the data tells the same story. These are Jupiter's Australian numbers, compiled last year by Drew Ianni. His US data showed the same patterns. So does the data from their biggest rival, Forrester Research. This is some of the most trustworthy data available. This data is also extremely stable. The chart looked much the same three years ago. Drew Ianni has joked that his colleagues in the novelty-driven world of Internet research business give him a hard time for delivering the same usage pattern findings year after year.
The popular media would like you to believe that the Net is hip, happening and cool, that it works on Internet time and constant innovation. That's the hype. In fact, we have a medium where usage patterns are well established and not changing very fast. Usage is centred on quite prosaic tasks. People use the Net to get stuff done. The Net is a utility medium. In fact it's about as cool as the phone book.
But what about Internet entertainment and art? Well, a few sites are disseminating art - mostly sites that celebrate the arts of the essay and the poem. But they're exceptions. Entertainment and art have never even made it onto the Jupiter charts. They haven't even grown as bandwidth slowly rises and the Internet reaches ever more consumers. The current patterns aren't changing. And they won't deliver the Entertainment Internet or the Visual Arts Internet. That will take a sharp change in currently stable consumer behaviour. And today's most hyped technology, streaming video, needs not just changes in consumer behaviour, but technological breakthroughs and economic transformations as well. If anyone in the room is contemplating using streaming video, then I simply appeal to you - don't do it unless you know exactly what you're getting into. The deeper reality is that the Net doesn't do emotional experiences well. Even the calculatedly touching and simple Cathy Freeman TV ads we saw during the Sydney Olympics couldn't be made to work on the Web. The technology keeps getting in the way. It's not just technological glitches that much up the experience. The very nature of the PC-presented Internet stops you from simply sitting back and immersing yourself in an emotional experience. What we've got for the near-term future is the Get-Stuff-Done Internet. The stuff that gets done, of course, may include looking up an online reference such as artandculture.com, or downloading the annual report of an arts organisation, or checking a phone number in an online organisational phone directory. The Internet may not be the right setting for most art, but it can be very useful to the people who make and manage art and cultural activities. And I should point out the small group of online artistic endeavours that have cleverly exploited the Web's ability to do cheap tecxt publishing - succeses such as the Fray site and Australia's own Jacket, run by the poet John Tranter.
So much for what you can do. What will it cost you to do it? The short answer is that Internet development can be a fine investment if you do it right. But most people do it wrong. And if you do it wrong, it can rob you of vast amounts of money and management time. And few arts and cultural organisations have that much money to throw away, last time I checked.
First, a horrible truth: no-one has ever even found a convincing way to demonstrate the pay-offs from IT investment generally. Paul Strassman, a former chief information officer of Xerox, even wrote a book - "The Squandered Computer" - attacking IT professionals, consultants and academics who accepted strings of anecdotes as proof that IT spending was boosting productivity. Strassman wants these people to find rigorous ways to assess IT's contribution. Obviously much IT spending is useful - but which spending, and why? Strassman's 1997 book "The Squandered Computer" sported on its cover this scatter-plot graph, suggesting a clear non-relationship between corporate IT spending and investor returns. Macroeconomists have also searched for the link between IT spending and productivity. They haven't found it either. The Nobel Prize-winning economist Robert Solow famously declared that "we see computers everywhere but in the productivity statistics".
The lack of big-picture evidence for IT investment pay-offs matches neatly with the known difficulty of software project management. Software project management has a 40-year history of costing more than you thought. For example, there's a maxim in software engineering called the 90-90 rule: "The first 90% of the code accounts for the first 90% of the development time. The remaining 10% of the code accounts for the other 90% of the development time." The really bad news is that based on the available research, the 90-90 rule understates the problem. Somewhere between 20 and 35 per cent of software projects get cancelled. And so often do software projects defy management, that the average completed project actually takes 220 per cent of estimated time (McConnell). According to Paul Strassman, only sixteen percent of projects show successful completion on time and on budget. While the project runs, you pay developers the best part of $1000 a day. And on a Web site project, you're also paying people to somehow integrate the work of the developers with that of designers, marketers and various content suppliers. Once the thing is built, most people relax. They think the jobs over,. And that means they forgot to include maintenance costs. Those costs will go on year after year, at an estimated 15 to 50 per cent of the initial build cost. Most IT consultants and professionals don't tell you this stuff. It's not in their interests to do so. If they did, managers would approach IT projects with much more caution. "Most IT departments don't deliver what they promise," says Alan Pelz-Sharpe, a senior consultant with the UK-based IT consultancy Ovum. "The history of IT is a history of failure." Beating that history is very, very tough.
And most online projects have not beaten history. Most commercial online projects of the past five years have destroyed capital. All those billions poured into dot-coms, for barely any return. What goes wrong? Here I want to make an acknowledgement: I owe a lot to the consultant Chris Thomas, a start-up veteran and former colleague of mine at eChoice. We've discussed these issues frequently over two and a half years. The normal trap that an organisation falls into as it heads into an online project is the Sort-Of Trap. It sort-of knows what it wants to do, because it has some good people with business experience. They have no idea at all about implementation, but they want to implement. So they hire people and firms who have built software and systems. Those parties have their own interests. They want to bolster their CVs. The individual developers want to work with bleeding-edge technology. The more pragmatic firms want to re-apply previous solutions. They'll say "what do you want?" They'll say "give us a specification". And the client will respond with generalities, because they haven't thought out the detail of the project. And with that, they will cede all control to their technological hired guns. They'll abrogate responsibility. The hired guns will eventually come back with a solution and say "is this what you meant?" And at this stage the organisation's people say "no, we want that". And they end up with something with poor architecture and poor documentation, which can't be expanded over time or even easily maintained. And no-one wants to manage the results, not just because measurement is hard but because they're scared of what they'll find. And it's management's fault, because they ceded control. They refused responsibility for their own destiny. They should have defined their objectives for themselves, and then iterated them with their development team, building up a relationship of trust as they went. That's how you find the best solution. Strassman calls this process "alignment".
Now let's look at a specific example - Scape, the online entertainment joint venture between Village and Channel Ten. Scape is interesting because it appears to have made all the available mistakes.The big ones first:
Now ...
Now ... Quarter 1 2001 "McKinsey Quarterly" magazine, under the heading "E-performance: The path to rational exuberance" presents three simple principles based on a study of successful and failing Internet businesses. McKinsey fell for Internet hype big-time during the dot-com boom, but these principles capture current best practice quite nicely. Principle 1: Find a well-defined group of users who will get value from your system early on, rather than requiring huge economies of scale and synergies. If you're a for-profit, that means early customers. Avoid the big up-front investments that Net-Gain-style McKinseyites used to advocate. In other words, adopt the Microsoft model rather than the Netscape one. Having users is more important than you can imagine. Build a working, useable prototype - a "pathfinder" project that will teach you enough to build the grown up project later. The customers will tell you what they want and don't want. Principle 2: Stick to your niche. If you sell CDs, diversifying into perfume is probably a mistake. Businesses have been told this ever since the 1970s conglomerate era ended, and apparently it works for Internet businesses too. Principle 3: Avoid bleeding-edge technology. The top performers spend their money on flawless delivery of the basic service or product, rather than on the ultra-cool new software that will have to be fixed in a year's time. I'll add a few principles of my own. Test for usefulness, usability, and use. Abandon your wounded - be prepared to compromise to survive. Use the 80-20 rule and kill unnecessary "nice-to-haves" (if they turn out to be must-haves, you can always build them later).
And never abrogate responsibility. Understand what's achievable. Decide what you want to achieve, what problem you want to fix, what opportunity you want to exploit. Then work with the experts to get what you want and need.
|
